Langmuir 1996,11, 3153-3160
3153
Langmuir-Blodgett Films of 3-AlkylpyrrolesStudied by Scanning Tunneling Microscopy W. M. Sigmund,?T. S. Bailey,? M. Hara,* H. Sasabe,*W. Knoll,*and R. S. Duran*l? Department of Chemistry and Center for Macromolecular Science and Engineering, University of Florida, Gainesville, Florida 3261 1, and Frontier Research Program, The Institute of Physical and Chemical Research (RIKEN), Wako, Saitama 351-01, Japan Received November 29, 1993. I n Final Form: April 6, 1995@ By use of the Langmuir-Blodgett (LB)techniquemono-and multilayersof monomers and their polymers were built up on conductive solid substrates for studies with scanning tunneling microscopy (STM). Monomers, mixed monomers, and polymer monolayers of 3-alkylpyrroles were transferred onto freshly cleaved molybdenum disulfide(MoSz)and highly oriented pyrolytic graphite (HOPG).High-qualityimages of the mono- and multilayers could be observed by STM after annealing. Monolayers of monomer were highly ordered showing orientational and positional correlationof alkyl side chains. Mixtures of pyrroles with different side chain lengths showed phase separation. Polymer monolayers showed high positional and orientationalorder and increased sidechain packing density with sidechain orientationmore orthogonal to the surface. Polymer multilayers showed disorder, revealing connectivity and curvature of polymer chains. From the structures observed on solid substrates (gas-solid interface) conclusions are drawn for the polymerization process and the packing of molecules on the Langmuir trough (at the gas-liquid interface).
Introduction Restricting monomer molecules to monolayers at planar surfaces during polymerization yields ordered anisotropic polymer thin films. Additionally, as the resulting polymer chains are prohibited from overlapping, the monolayer formed is highly two-dimensional (2-D). This type of ordered polymerization can be done at solid-liquid, liquid-liquid, or gas-liquid interfa~es.l-~ Polymerizations at solid interfaces, however, are limited by the slow lateral diffusion of monomer and even slower diffusion of oligomer and polymer, which finally limits the size of the molecule. For a fluid interface, diffusion is much faster, and when the applied surface pressure is maintained constant, limited conversion due to lateral shrinkage is not a problem. Although orientation at the aqueous surface greatly reduces the number of conformations the chains can assume, many possibilities still exist. These are described in the model of a “true self-avoiding walk, that was originally used for modeling the adsorption of polymers onto solid substrates without allowing overlapping of chains. Since the adsorption process is analogous to chain growth during polymerization at the interface, this model can directly be applied for the description of the conformations of a growing polymer chaina4Besides the random walk of the chain growth at the interface, other factors contribute to the packing of the molecules including phase separation of monomer, oligomer, and polymer, viscosity of the film, and polymer backbone stiffness, which all can be combined into kinetic and thermodynamic contributions. No direct experimental observations have been reported to prove the two-dimensional order of LB polymer films on water. Furthermore, although many types of 2-D University of Florida. The Institute of Physical and Chemical Research (RIKEN). Abstract published in Advance A C S Abstracts, June 15,1995. (1)Gaines, G. L. Insoluble Monolayers at Liquid-Gas Interfaces; Wiley: New York, 1966; pp 321-2. (2)Sigmund, W. M.; Duran, R. S. Macromolecules 1993,26,2616. (3)Ulman, A.A n Introduction to Ultrathin Organic Films: From Langmuir-Blodgett to Self-Assembly;AcademicPress: New York, 1991; @
pp 176-185.
(4)Bulgadev, S.;Obukov, S. Sou. Phys. JETP 1984,59,1140.
polymerizations at gas-liquid interfaces (Langmuir polymerizations) have been reported in the literature, no imaging data at the molecular scale exists describing the conformation of the resulting polymer chains, the packing of the chains, and the organization of the domains. By introducing an electron-rich monomer, whose electrons are easily removed as they are close to the Fermi level, and a conductive polymer, we report on scanning tunneling microscopy (STM)investigations of 2-D polymerized monolayers. The goal is to elucidate the packing of the monomer and the main chain conformation of its polymer. 3-Decylpyrroleand 3-hexadecylpyrrole and its polymer were used for these investigationsas they form monolayers at the aqueous interface and can be transferred to solid substrate^.^-^ We will compare the ordering observed in monolayers of monomer, mixtures of monomers, polymer, and polymer multilayers, all of which were transferred to molybdenum disulfide (MoSz)or highly oriented pyrolytic graphite (HOPG) substrates and then annealed. MoS2 was chosen because the anchoring structures on it tend to reflect the intrinsic ordering of adsorbates because ofits 6-fold It also provides favorable STM contrast conditions that give bright images of molecules containing long alkyl chains and is molecularly smooth without dangling bonds. HOPG was chosen as a second substrate, because it is well studied and gives a different STM contrast for alkyl chains and a different anchoring structure than MoS2. STM has proven to be a powerful tool for studies of the conformation of low molecular weight adsorbates on solid substrates. Furthermore, it has been used to study packing in transferred Langmuir-Blodgett (LB) monol a y e r ~however, ;~ limited molecular information could be ( 5 ) Duran, R. S.; Zhou, H. C.Polymer 1992,33,4019.
(6)Sigmund,W.M.; Marestin, C.; Keil, S.; Zhou, H.; Elsabee, M. Z.; Riihe, J.; Wegner, G.; Duran, R. S. Submitted for publication in
Macromolecules. (7)Rahman, A. K. M.; Samuelson, L.; Minehan, D.; Clough, S.; Tripathy, S.;Inagaki, T.; Yang, X. Q.;Skotheim, A. T.; Okamoto, Y. Synth. Met. 1989,28,C237. ( 8 ) Iwakabe, Y.; et al., Jpn. J . Appl. Phys. 1991,30, 2542. (9)Smith, D. P.E.; et al. Proc. NatLAcad. Sci. U S A . 1987,84,969.
0743-746319512411-3I53$09.OQ/O 0 1995 American Chemical Society
3154 Langmuir, Vol. 11, No. 8, 1995
Sigmund et al.
,
.,
L
*
N H
Figure 1. Chemical structures of the monomer (A) 3-hexadecylpyrrole and its polymer (B) poly(3-hexadecylpyrrole).
gathered. This limited information has been attributed to the poor conductivity of the LB layer, a direct consequence of molecular packing orthogonal to the substrate plane. As a result, atomic force microscopy has been used more often to image LB films.lOJ1In our case, the intrinsic conductivity of the polypyrrole backbone a n d the pyrrole ring in the monomer should change the contrast a n d allow us to image both monolayers and multilayers by STM more successfully than conventional LB materials. Experimental Procedures and Chemicals Synthesis of the monomers, 3-decylpyrrole and 3-hexadecylpyrrole, was based on previously published procedures.12 Both compounds were found to be pure by 'H-and 13C-NMR (XL-200NMR, Varian), TLC and FTIR. Elemental analysis of 3-hexadecylpyrrolewas as follows: Calcd: C, 82.98; H, 12.19; N, 4.84. Found: C, 82.62; H, 12.74; N, 4.76. Monolayers of the monomers and the polymer were studied with a commercially available computer-controlled film balance (KSV-5000,KSV Instruments, Finland). The interfacial pressure was monitored by suspending a platinum plate through the interface and applying the Wilhelmy method with on-line data acquisitionthrough a surface balance. Monolayerswere obtained by spreading a solution of the monomer dissolved in chloroform (ACS Spectro Grade, Eastman) at concentrations of approximately 0.5 mg/mL. The balance was zeroed before addition. Isotherms of surface pressure versus mean molecular area per repeat unit (Mma)were measured at 23 "C on Milli-Qwater with a resistance of greater than 18 MQkm for the monomers. For polymerization and studies of the polymer, subphase solutions were made with ACS reagent grade chemicals and Milli-Qwater. Subphase oxidizing agents were ferric chloride (0.03 M) or ammonium peroxydisulfate (0.03 MI. After polymerization, materials collected from the trough surface for further analysis were washed with water dissolved in chloroform and after evaporation dried in a desiccator. For STM studies the polymerized monolayerswere transferred directly after the two-dimensional polymerization applying Langmuir-Blodgett (vertical dipping) or Langmuir-Schafer technique (horizontal dipping). The layers were deposited on freshly cleaved MoS2 (Hori Mineral Co.) and HOPG (Advanced Ceramics), with a KSV minitrough. Conditions for transfer were room temperature (21 "C), substrate dipping speed 5-10 mm/ min with a surface pressure for transfer of 20 mN/m. Monolayer samples were prepared by one upstroke of the substrate through the interface, placing the hydrophilic pyrrole rings or the polypyrrole backbone at the substrate surface, as well as by one downstroke and a high-speed upstroke with negligible transfer observed, putting the alkyl chains onto the solid surface. Multilayer samples were prepared by alternating up and down strokes through the interface. The horizontal transfer was done by lowering the substrate with a micrometer screw. With this technique only monolayer samples were prepared. (10) Zasadzinski, J. A. N.; et al. Biophys. J. 1991,59,755. (11)Schwartz, D. K.; Garnaes, J.;Viswanathan, R.; Zasadzinski, D. P. E. Science 1992,257,508. (12)Riihe, J.;Ezquerra, T.A.; Wegner, G. Makromol. Chem. Rapid Commun. 1989,5,1.
10 20 30 40 Mean Molecular Area per Molecule
0
(AZ) Figure 2. Surface pressure vs mean molecular area during compression for 3-hexadecylpyrrole (A) on water and poly(3hexadecylpyrrole) (B) on 0.03 M ammonium peroxydisulfate subphase solution: temperature 23 "C; surface pressure for polymerization 10 mN/m: compression barrier speed 20 mm/ min. STM pictures were obtained on a NanoScope I1 in the constant current mode in air. Tip bias voltages ranged from 1100 to 1600 mV for the monomer and -765 to -1980 mV for the polymer. Tunneling currents ranged from 0.09 to 0.37 nA. Tips (Pt-Ir) were bought from Digital Instruments, CA, and recut with scissors as needed. To achieve high-resolution images, samples were annealed for at least 12 h above the melting temperature of the monomer before investigation. Then, they were allowed to cool down at a rate of 5 "C per 15 min with a pause of 45 min close above the melting temperature ofthe monomer. Deposition and STM imaging were repeated several times; thus the images presented in the subsequent text are representative. To modify the different anchoring structures for some polymer monolayers, a drop of chloroform was spread with a Hamilton syringe onto the substrate while scanning.
Results and Discussion Monomerand Polymer at a Gas-Liquid Interface. The isotherm of 3-hexadecylpyrrole is depicted in Figure 2, curve A. On a p u r e w a t e r surface, the onset of the isotherm is at a M m a per molecule of 26.5 Az. Its collapse pressure is 6 3 mN/m. Before collapse, a highly viscous film is observed that can move the Wilhelmy plate when working o n a single barrier trough. By addition of ammonium sulfate to the subphase, no change in t h e isotherm can be observed u p to concentrations of 0.03 M. The s a m e has been reported for ferric chloride in concentrations u p to 0.01 M.' The observedincreasein surface a r e a can be interpreted as a phase transition from a liquid condensed to a liquid expanded layer. The s a m e type of phase transitions have been reported by Herrir1gt0n.l~ ~
~~
(13)Herrington, T. M.; Aston, M. S. J. Colloid Interface Sci. 1991, 141, 50.
STM Study
of 3-Alkylpyrroles
Langmuir, Vol. 11, No. 8, 1995 3155
Figure 3. Two STM images of 3-hexadecylpyrrole monomer on MoS2 scanned in constant current mode at room temperature in air: dimensions of the pictures 15 nm x 15 nm; tip bias voltage 1320 mV; current 0.31 nA. STM samples were prepared by vertical transfer of preformed monolayers to MoSz substrates. Monolayers transferred vertically in upstroke from pure water subphases onto fresh cleaved MoS2. The images are of two different regions on the same substrate, the different color scales were chosen to guide the eye.
The amphiphilic molecule is thought to place itself accordingto its polarity and to the polarity of the interface, i.e., the more polar pyrrole group is in the water whereas the alkyl chain points into the air. From the tangent of the isotherm the mean molecular area of one mqlecule at zero surface pressure can be measured as 22.6 A2. With a few simple assumptions the average arrangement of the molecules at the surface can be estimated. First, the molecules are considered to be rigid rods. Second, they form a dense packing on a surface. Third, the rods have a diameter of 0.36 nm. An average tilt angle to the surface of 72.2" can then be calculated. A 2-D polymer can be formed by reacting the monomer 3-hexadecylpyrrole with an oxidizing agent in the subphase. During polymerizationcovalent bonds are formed between monomers, therefore bringing them closer together. This can be observed macroscopically by the decrease in surface area, when polymerization is done at constant surface pressure. After polymerization the monolayer can be expanded and recompressed; a recompression isotherm is shown in Figure 2, curve B. The collapse pressure for the polymer film is lower compared to the monomer and the tightly collapsed film is dark colored with molecular weight averages of 3500 to 6000 measured by GPC using polystyrene standards. The point of onset in the first compression curve of the polymer is at a larger Mma than for the monomer isotherm. This seems to be contradictory since during polymerization, the Mma of the monolayer decreased. This apparent contradiction can be explained by the following: First, the monomer isotherm is dependent on the salt concentration and type of salt in the subphase, as discussed above. Second,after polymerizationthe monolayerwill rearrange and demix on decompression,simultaneously allowingthe pyrrole group to lie more flat on the surface. Hysteresis experiments indicate that this is the case since on repeated compression-decompression cycles the isotherm reaches a constant value that is equivalent to the value directly after polymerization. Transfer onto substrates is possible with transfer ratios being about unity.6 For more experimental details about
the Langmuir polymerization of 3-hexadecylpyrrole,we refer t o the existing l i t e r a t ~ r e . ~ ? ~ > ~ J ~ Monomer at a Gas-Solid Interface. Images of a 3-hexadecylpyrrolemonomer monolayer at the MoSz-air interface obtained by STM are shown in Figure 3. The dimensions of the pictures are 15 nm by 15 nm. Rodlike structures are observed showing high orientational and positional order. The rods form rows with a row-to-row distance of about 1.5 nm and are tilted at an angle of 40" to the rows. The measured rod length is 1.7 f 0.1 nm, and the thickness is about 0.36 nm. Additionally, it is clear from the image that the rods are interdigitated, i.e., the end of one rod lies in between two other rods of the next row. The rodlike structures can be correlated with monomer molecules, each rod corresponding to one molecule. By use of the simple model with cylindrical rodlike molecules (molecularwidth 0.36 nm and length 2.06 nm) in a dense packed structure, an angle to the surface of 12.2" can be calculated. The width of the rods of 0.36 nm can be deduced from the STM measurement. The assumption of a stiff rodlike all-trans alkyl chain may be expected for a hexadecyl chain which is likely to crystallize. Furthermore, the images suggest that this might be so. This result implies that at an angle of 12.2" only 17.4 nm of the molecule is visible as demonstrated in Figure 4B. By use of bond distances, however, the calculated length of 3-hexadecylpyrroleis about 2.06 nm; therefore, about 0.32 nm of the molecule is covered and cannot be seen by the STM tip. To distinguish which part of the molecule it is, the hexadecyl chain or the pyrrole ring, experimental parameters were changed while scanning the same area. An electron-rich pyrrole ring, where electrons are closer to the Fermi level than in alkyl chains, is expected t o show different tunneling conditions than alkyl chains. If the pyrrole group was in the visible part of the molecule, a change in the tunneling conditions should show a different contrast for some part ofthe rodlike (14)Hong, K.; Rosner, R. B.; Rubner, M. F. Chem. Muter. 1990,2, 82-88.
Sigmund et al.
3156 Langmuir, Vol. 11, No. 8,1995
Figure 4. Pictorial model calculated from the observed packing of monolayers of 3-hexadecylpyrrole: (A) the air-water interface; (B)the MoSp-air interface. The rigid rods have a length of 2.06 nm and a diameter of 0.36 nm.
structure. In fact, an overall change can be observed, but no features in the rodlike structure could be identified that could be correlated with the ring. Therefore, we conclude that the pyrrole ring is not visible and must be attached to the substrate. Additionally, the observed packing distances do not allow the ring to lie flat on the substrate; the ring must be standing. The attachment of the ring is also expected from the transfer method; an upstroke places the hydrophilic pyrrole ring on the substrate surface and the alkyl chains in the air. Annealing above the melting temperature of the bulk might cause a complete reorganization, but it is also themodynamically expected that the more polar ring should be adsorbed on MoS2 and the less polar alkyl chains a t the air surface form a low-energy surface.15 Comparison of 3-Hexadecylpyrrole Monolayers at a Liquid-Gas and a Solid-Gas Interface. The mean molecular area on water is 22.6 A2, whereas that on MoS2 is 57 Hi2. The increase in surface area per molecule of 34.4 A2 implies that during annealing, about half the molecules must be expelled from the monolayer. Some differencesin the molecular arrangementof the monolayer at the liquid-gas and solid-gas interface are shown in Figure 4. The angles formed by the alkyl chain on water and MoS2 are calculated to be 72.2" and 12.2",respectively. It should be mentioned that no surface pressure is externally applied to the monolayer at the solid surface, whereas the freedom of the molecules in the layer at the liquid interface is mechanically restricted by a barrier. No images could be obtained when samples were scanned directly &r deposition,without annealing. Since STM, as we used it, did not allow the detection of very irregular structures on a molecular scale, we assume that the monolayer directly after deposition is not well ordered. The layer had to be annealed above the bulk melting point to allow for an ordered thermodynamically stable monolayer structure to form. Therefore, as used here, LB transfer is only a tool that allows for convenient and relatively uniform initial film thickness in the nanometric scale, not to establish the nature of the lateral packing of the molecules. It should also be noted that melting and diffision of a few micrograms of 3-hexadecylpyrrole on MoS2, a common sample preparation method for STM observation of other low molecular weight organics, did not give monolayers which were observable by STM.16 (15) Koopman, D. C.; Gupta, S.; Ballamudi, G. B.; Westermann-Clark, G. B.; Bitsanis, I. A. Submitted for publication in J. C k m . Phys.
Mixtures of 3-Decylpyrroleand 3-Hexadecylpp role. Scanning pure 3-decylpyrrole monolayers always resulted in images of the clean MoSz surface. Bulk 3-decylpyrroleis a liquid a t room temperature, its melting temperature being 16.5 "C. Therefore, it is expected that the compound is moved easily by the tip while scanning the surface. Scanning mixtures of 3-decylpyrrole and 3-hexadecylpyrrole gave the images shown in Figure 5. A mixture of 95% 3-hexadecylpyrrole in 5% 3-decylpyrrole was transferred in an upstroke from the water surface onto MoS2. The STM images in parts A and B of Figure 5 show a region of 13.5 nm2. Clearly separated domains of different rodlike structures can be recognized in all images. Again, we assume that each of these rods correspondsto one molecule. Since mixtures were scanned, it is not clear which rod corresponds to each monomer. The length of the rods observed by the STM tip ranges from 1.2 to 1.9 nm. Since we cannot directly correlate the rods with the type of molecule, no angles of attachment with respect to the substrate layer plane can be calculated. The different measured rod lengths may correspond to the same type of molecule attached a t a different angle. They also may correspond to different monomers attached at different angles. Generally, most of the rods are assumed to correspond to the 3-hexadecylpyrrole,since its spreading ratio was 95%. The ratio is very likely to shift to higher values for the hexadecylpyrrole because the decylpyrrole is slightly soluble in water and will diffuse to a certain extent into the subphase before transfer. Additionally, the decylsubstituted compound is a liquid in the bulk and can be moved easily by the tip a t the surface. The structures observed by the tip, therefore, have to be thermodynamically stable; i.e., only smaller domains of the shorter alkyl chain compound that are stabilized by domains of the hexadecyl compound or mixtures of the compounds or domains of different mixtures will be visible. The structures that can be found on these samples are shown in Figure 5. Figure 5A shows a region where formation of domains occurs. These domains consist of mixtures with different amounts of the hexadecyl and decyl compound. We can conclude this because we find rodlike structures ranging from 1.2 to 1.9 nm in all domains. 3-Decylpyrrole lying completely flat would only allow a length of 1.4 nm and a pure 3-hexadecylpyrrole would show the packing seen in Figure 3. Additionally, there is not enough 3-hexadecylpyrrole transferred to cover the surface with such a dense packed monolayer. Figure 5B shows a region of the same sample with two domains. On the left side of the image, a domain with rows of molecules having positional and orientational order is seen. On going to the center, the positional order is reduced, while the orientationalorder is retained resulting in a 2-D nematic-like ordering. No interdigitation is observed either in the region with rows or in the 2-D nematic phase region. The image in Figure 5C shows a region of the size 18.75 nm that is ordered in rows with a row-to-row distance of 0.75 nm and no interdigitation of the rods. We expect this region more likely to be pure 3-decylpyrrole because of the different packing structure. The image in Figure 5D was scanned in the same region with a higher magnification. The image is 6.75 nm2. The rods that can be seen have a length of 1.4 nm and form an angle of 32" to the rows. Ifwe assume that this is a pure 3-decylpyrrole (16)Hara, M.;et al. Nature 1990,344,228.
STM Study of 3-Alkylpyrroles
Langmuir, Vol. 11, No. 8, 1995 3157
4
Figure 5. STM images of a mixture of 3-decylpyrrole and 3-hexadecylpyrrole monomer monolayers on MoS2: (A and B) 13.5 nm x 13.5 nm, tip bias voltage -1364 mV, current 0.12 nA; (C)18.75 nm x 18.75 nm, tip bias voltage -1364 mV, current 0.13 nA; (D) 6.75 nm x 6.75 nm, tip bias voltage -1364 mV, current 0.13 nA.
region and apply the rigid-rod model, we can conclude that the molecules must lie completely flat on the surface because the length of the rods equals the length of 3-decylpyrrole. This implies that the dark area at the end of the rod that indicates a thinner overlayer or good conductivity is very likely to be the pyrrole ring. The length of this darker area equals 0.32 nm, which is exactly the amount that is not visible in the packing structure of pure 3-hexadecylpyrrole. Overall,we can concludethe followingfrom the mixtures of monomer: Different types of packing structures are possible. Packing is possible in domains having positional and orientational order with and without interdigitation, mixed domains tending toward structures having less positional order, only keeping one major axis of order, Le., 2-D nematic phase formation. While quantitative interpretation of surface compositionsis not possible from these data, the packing of the molecules seems to be highly influenced by the surface structure of the substrate. The minimization of surface energy for the monolayer that leads to coverage of the pyrrole group by the alkyl chains in the hexadecyl compound is overruled by the enthalpic adsorption onto the sites along one of the major axis of the
substrate surface for the decyl compound,which is a liquid in the bulk. Polymer Monolayers at Gas-Solid Interfaces. Poly(3-hexadecylpyrrole) was transferred via vertical depositionon HOPG (Figure 6a),vertically in one upstroke or one downstroke on MoS2 (parts b and c of Figure 6, respectively)and horizontally on MoS2. Furthermore, the horizontally deposited monolayer was solubilized with chloroform while scanning and higher resolution images (Figure 6d-f) could be obtained after evaporation. Figure 6a shows a STM image of poly(3-hexadecylpyrrole) on HOPG. Rodlike structures can be seen that form an ordered structure in rows, the row-to-rowdistance being 0.59 nm. T$e rods form an angle of 27" with the rows. A Mma of 32 A2can be calculated that compares well to the monolayer at the liquid interface. Scanning a vertically-dipped polymer monolayer on MoS2 gives the image shown in Figure 6b. Rodlike structures organized in rows with a row-to-row distance of 0.9 nm and a rod length of 1.4 nm leading to an angle of 40" and a Mma per repeat unit of 49 A2 can be seen. The same structure is observed for polymer monolayers
3158 Langmuir, Vol. 11, No. 8, 1995
Sigmund et al.
(b)
Figure 6. STM images of poly(3-hexadecylpyrrole) monolayers in constant current mode in air. Substrate, type of deposition, and dimensions for images are as follows: (a)HOPG, vertical upstroke, size 10 nm x 10 nm, tip bias voltage -765 mV, current 0.37 nA; (b) MoS2, vertical upstroke, size 7 nm x 7 nm, tip bias voltage -1400 mV, current 0.2 nA; (c) MoS2, vertical downstroke, size 10 nm x 10 nm, tip bias voltage 1599 mV, current 0.13 nA; (d) MoS2, horizontal deposition, CHC13, size 6 nm x 6 nm, tip bias voltage -1362 mV, current 0.23 nA; (e) MoS2, horizontal deposition, CHC13, size 6 nm x 6 nm, tip bias voltage -1362 mV, current 0.23 nA; (0MoS2, horizontal deposition, CHC13, size 6 nm x 6 nm, tip bias voltage -1362 mV, current 0.14 nA.
that were transferred with one downstroke and then annealed (Figure 6c). The annealing, therefore, appears to change the orientation of the deposited film, placing the polymer backbone at the substrate surface. Horizontal dipping and annealing lead to comparable structures as the vertical deposition. However, after a drop of chloroformis spread while scanning, a new packing with lower densioty can be observed. A Mma per repeat unit of about 60 A2 resulted. Furthermore, faint streaks at the end of each rod and in the row direction may be due to the presence of the polymer backbone. In any case, the chloroform appeared to “swell” the polymer monolayer, attaching the backbone to a major axis of orientation of the substrate. The polymer ordering was observed on MoS2 over a wide range of scanning condition: tip bias voltages both positive and negative between 700 and 2200 mV and tunneling currents between 0.07 and 1.0 nA. The above behavior is in contrast to the monomer, where increasing the tunneling current or changing the tip bias voltage toward zero easily resulted in molecules being scraped off by the tip, revealing the underlying MoS2 lattice. No features in any of the polymer images could be associated with counterions; while some counterions are expected to be
associated with the transferred film, they should be relatively mobile and not highly ordered, therefore not easily visible by STM. Independent of substrate and transfer protocol, the following features can be observed which provide useful information for comparisonwith the monomer monolayers. Similar rodlike structures are observed that display orientational alignment perpendicular to the scanning direction. Each rodlike structure may correspond to a single alkyl side chain along the polymer backbone. The backbone is not clearly visible, perhaps only becoming partially visible upon swelling by chloroform. The side chain packing density in the polymer monolayer on graphite is greater. This suggests that the anchoring interactions on graphite are weaker than those on MoS2, and polymer-polymer interactions, much like those at high applied surface pressure on the water surface, dominate the anchoring structure formation on graphite. The side chain density on MoS2 is related with two structural distances. The first is the chemically defined repeat distance determined by both the chemical bond lengths and the configuration of the underlying polymer backbone. The second involves the proximity of one polymer chain to another.
Langmuir, Vol. 11, No. 8, 1995 3159
STM Study of 3-Alkylpyrroles
Substrate
Figure 7. Pictorial model of the polymer monolayer structure observed at the MoSz-air interface without swelling by CHC1:j. Alignment with double spacing of the MoS2.
On MoS2, the alkyl side chain ordering observed in the polymer image can be explained by an all-transoidal conformation of the (polymerall-transoidal conformation of a 2,5 head-to-tail condensation) polypyrrole backbone with respect to the nitrogen, such that each polymer chain arranges its alkyl side chains into two parallel rows. In this conformation,the distance between these two parallel rows of alkyl side chains corresponds to nearly twice the MoS2 lattice spacing. Polymer main chains are then aligned in a parallel fashion suggestive of the overall conformationdepicted in Figure 7. Ifthe distance between the polymer main chains is also twice the MoSn lattice spacing, then this conformationvery accurately explains the uniform structure and the two directions of row alignment observed in the STM images. We suppose there are several reasons that lead to the observed polymer anchoring structure on MoS2: the backbone oriented parallel to a major axis of MoS2 reduces the interfacial energy; the backbone at the MoS2 and side chains extended toward the air reduce the surface energy. The parallel alignment of adjacent polymer chains and that of the alkyl side chains allows for the best packing density at the surface. Similar anchoringinduced ordering has been observed for systems of low molecular weight rodlike molecules16 such as cyanobiphenyl containing liquid crystals. While the images obtained are believed to represent the intrinsic order ofthe layers, it is not known to what extent the act of scanning influences the ordering. It would be interesting to be able to scan the sample at arbitrary angles to further investigate this point.
Polymer Multilayers and Molecule-SubstrateInteractions. Figure 8 shows STM images of a polymer multilayer. Rodlike structures are seen that may correspond to the alkyl side chains. The alkyl side chains no longer show very high orientational correlation and they do not uniformly orient perpendicular to the scanning direction. Generally, an increase in disorder is observed. The rows of alkyl side chains are existing but curvature is visible, which gives evidence of connectivity between them. Each polymer chain still shows positional row correlation but the chains now appear to “drape”themselves over defects in the layers below and to show curvature in the layer plane. Interrow distances also show local variations not observed in the monolayer images. The multilayer structure observed may be described as follows: The innermost layer against the MoS2 surface is very highly ordered with the polymer backbone oriented directly against the substrate surface. Subsequent overlayers are much more disordered, and more closely reflect the polymer backbone conformation after polymerization at the gas-liquid interface. The long alkyl chains of the first and second layer can be compared to as solvent for the overlayer, therefore also changing the surface characteristic from a solid to a more liquid-like one. The effects of the annealing process in monolayers of both the monomer and polymer can be compared. Both systems required annealing in order to achieve very high molecular resolution by STM. This emphasizes the importance of molecule- substrate interactions in defining surface order. However, the degree which annealing affected the two systems was very different. As discussed earlier, the monomer showed a tendency to be easily reordered from its original packing in the LB layer by extrusion of some molecules from the surface. For the polymer, respacing and realignment of the polypyrrole backbone with a major axis of the MoS2 appears to have occurred,but generallythe layer showed a greater stability associated with higher molecular weight molecules and, therefore, more anchoring sites.
Figure 8. STM images of polymer multilayers of poly(3-hexadecylpyrrole): Dimensions of images, 10 nm x 10 nm; tip bias voltage - 1980 mV; current 0.14 nA. The sample was prepared from a mixed monolayer containing 64%of the low molecular weight liquid crystal 8CB, used as a processing aid and to establish additional disorder at the surface. Note the polymerization was carried out prior to the addition of 8CB. The samples were prepared by vertical deposition at an applied pressure of 8 mN/m and were nominally three layers thick. The samples were annealed after transfer of all layers.
3160 Langmuir, Vol. 11, No. 8, 1995 Very ordered structures with both orientational and positional correlation can be seen in all monomer and polymer images. For the monomer, the annealing process allows the molecules to orient parallel to the surface, whereas for the polymer, it appears the backbone becomes aligned parallel with a major axis of the MoSz substrate and establishes positional order. Thus, both the positional order and chemical repeat distance define the side chain packingdensity. As an effect,the side chains cannot align parallel to the solid surface as in the monomer monolayers. They remain partially erect due to steric interactions with adjacent side chains. In both cases, a significant degree of ordering is induced. The polymer backbone in the outer layer of the multilayer system is not orientationally ordered by the MoSz. It can take up conformations much more like those intrinsic to the polymerization at the gas-liquid interface. The molecular conformation or contrast conditions allow one to observe even greater evidence of the connectivity related to the backbone. The long alkyl side chains of the underlying monolayers serve as solvent for the top layers. Therefore, the conformation of the top layers shows increasing disorder compared to the first layer, where order appears to be highly dependent on the moleculesubstrate interactions. The observed images on solid substrates suggest the following conclusions about the monolayers and the polymerization process at the air-liquid interface. The nematic-like domain formation by small oligomer molecules must compete with the tendency for a polymer chain to phase separate as a single unit or to fold back and forth across it~e1f.l~ The last two conformational possibilities demand a greater degree of polymerization. At the gasliquid interface, then, rapid domain formation by small oligomer molecules is possible and would likely favor conformations dominated by parallel alignment of the polymer chains, Thus, the fluidity of the gas-liquid (17)Onsager, J.Ann. N.Y. Acad. Sci. 1949,52, 627.
Sigmund et al. interface may amplify the rate dependence of the resulting polymer conformations compared to those observed at a solid surface.
Conclusions The Langmuir trough and the LB technique were used as tools for establishing quasi-two-dimensionalityduring polymerization and uniform film thickness, not to define local lateral order within a layer transferred to MoSz as has been shown. For mixtures of monomers phase separation with domain formation can be concluded. It also can be concluded that the ordering of the monomer on MoSz is caused by organization parallel to a major axis of the lattice. For the polymer, the ordering of the backbone along a major axis of the lattice forces the side chains into a different packing than in the monomer. The structures seen are regularly ordered giving evidence that the polymer formed is highly regular 2,5 head-to-tail and in an all-transoidal conformation. This suggests that during the polymerization process, the densely packed monomer monolayer mainly allows polymerization in the 2,5 position. Another important result of these investigations was that the introduction of a conductive polymer backbone did allow the successful imaging of both LB monolayers and multilayers. Such conducting polymers are candidates for a new class of adsorbates to flexiblymodify MoSz and other substrates. Thus, using conductive polymer films as underlayers to change lattice spacing or contrast conditions may prove useful for studying other systems such as alignment in liquid crystals or functional group effects in molecule-molecule interactions. Acknowledgment. The Deutsche Forschungsgemeinschaft and the Officeof Naval Research are acknowledged for support. LA930693F
